This is a post-peer-review, pre-copyedit version of an article published in Fungal ecology. The final authenticated version is available online at: 10.1016/j.funeco.2018.08.001
©2018 Licensed under the Creative Commons CC-BY-NC-ND 4.0
From leaf to continent: the multi-scale distribution of an invasive cryptic pathogen complex on oak
Marie-Laure Desprez-Loustau1*, Marie Massot1, Maude Toïgo1, Tania Fort1, Gulden Aday2, Johanna Boberg3, Uwe Braun4, Xavier Capdevielle1, Thomas Cech5, Anne Chandelier6, Petya Christova7, Tamara Corcobado8, Tugba Dogmus9, Cyril Dutech1, Olivier Fabreguettes1, Julie Faivre d'Arcier1, Andrin Gross1,10, Marilia Horta Jung11, Eugenia Iturrixa12, Thomas Jung11, Corina Junker13, Levente Kiss14, Kaloyan Kostov7, Asko Lehtijarvi15 , Aneta Lyubenova7, Benoit Marçais16, Jonas Oliva3,17, Funda Oskay18,Martin Pastirčák19, Katarína Pastirčáková20, Dominique Piou21, Gilles Saint-Jean1, Arnaud Sallafranque1, Slavtcho Slavov7, Jan Stenlid3, Venche Talgø22, Susumu Takamatsu23, Ayco JM Tack24
Some institutes below have a full address, others not
1 BIOGECO, INRA, Univ. Bordeaux, 33610 Cestas, France
*corresponding author : [email protected], tel. +33 (0)535 38 53 20
2 2 Süleyman Demirel University, Yenişarbademli Vocational School, Isparta, Turkey
3 Department of Forest Mycology and Plant Pathology, Swedish University of Agricultural Sciences (SLU), P.O. 7026, SE-750 07 Uppsala, Sweden
4 Martin-Luther-Universität, Institut für Biologie, Bereich Geobotanik, Herbarium, Neuwerk 21, 06099 Halle (Saale), Germany
5 Federal Research and Training Centre for Forests, Natural Hazards and Landscape (BFW), Department of Forest Protection, Vienna, Austria
6 Walloon Agricultural Research Centre, Belgium
7 AgroBioInstitute, 8 Dragan Tsankov blvd., 1164 Sofia, Bulgaria
8 Federal Research and Training Centre for Forests, Natural Hazards and Landscape (BFW), Vienna, Austria; Phytophthora Research Center, Mendel University, Brno, Czech Republic (present affiliation)
9 Süleyman Demirel University, Faculty of Forestry, Isparta, Turkey
10 Present address: Eidg. Forschungsanstalt WSL, Zürcherstrasse 111, 8903 Birmensdorf, Switzerland
11 Laboratory of Molecular Biotechnology and Phytopathology, Centre for Mediterranean Bioresources and Food, University of Algarve, 8005–130 Faro, Portugal - Phytophthora Research Centre, Mendel University, 613 00 Brno, Czech Republic (present affiliation)
12 Neiker, Granja Modelo de Arkaute, Spain
13 Julius Kühn-Institut, Messeweg 11/12, 38104 Braunschweig, Germany
14 Plant Protection Institute, Centre for Agricultural Research, Hungarian Academy of Sciences (MTA-ATK), H-1525 Budapest, P.O. Box 102, Hungary - Centre for Crop Health, University of Southern Queensland, Toowoomba, Qld 4350, Australia (present affiliation)
15 Bursa Technical University, Faculty of Forestry, Bursa, Turkey
16 IAM, INRA, Univ. Nancy, 54280 Seichamps, France
17 Dept. Crop and Forest Sciences, University of Lleida, Spain (present affiliation) 18 Çankırı Karatekin University, Faculty of Forestry, Çankırı, Turkey
19 Research Institute of Plant Production, National Agricultural and Food Centre, Bratislavská cesta 122, SK-92168 Piešťany, Slovakia
20 Branch for Woody Plants Biology, Institute of Forest Ecology, Slovak Academy of Sciences, Akademická 2, SK-94901 Nitra, Slovakia
21 Département Santé des Forêts, France
22 NIBIO - Norwegian Institute of Bioeconomy Research, Pb 115, NO-1431 Ås Norway
23 Graduate School of Bioresources, Mie University, 1577 Kurima-Machiya, Tsu 514- 8507, Japan
24 Department of Ecology, Environment and Plant Sciences, Stockholm University, SE- 106 91 Stockholm, Sweden
4 Authors contributions: MLDL and AJMT conceived and designed the research, contributed
to data acquisition, analyzed and interpreted the data and wrote the manuscript; MLDL organized the research; MM, MT, TF and ST performed molecular analyses and statistical analysis (MT) of data and contributed to the draft manuscript. GA, JB, UB, XC, TC, AC, PC, TCS, TD, CD, OF, JFA, AG, MHJ, TJ, EI, CJ, LK, KK, AL, AL, BM, JO, FO, MP, KP, DP, GSJ, AS, SS, JS and VT conducted sampling, especially at the European scale, and were involved in discussions and manuscript editing. All co-authors approved the final version.
Descriptors: cryptic invasion, cryptic species, Erysiphe, oak, plant pathogen, powdery mildew, Quercus, spatial distribution, species coexistence I would not include keywords that are present in the title
Abstract
The spatial distribution and niche differentiation of three closely related species (Erysiphe alphitoides, E. quercicola and E. hypophylla) causing oak powdery mildew was studied at scales ranging from the European continent, where they are invasive, to a single leaf. While E.
alphitoides was dominant at all scales, E. quercicola and E. hypophylla had restricted geographic, stand and leaf distributions. The large-scale distributions were likely explained by climatic factors and species environmental tolerances, with E. quercicola being more frequent in warmer climates and E. hypophylla in colder climates. The extensive sampling and molecular analyses revealed the cryptic invasion of E. quercicola in nine countries from which it was not yet recorded. The presence of the three species was also strongly affected by host factors, such as oak species and developmental stage. Segregation patterns between Erysiphe species were finally observed at the leaf scale, between and within leaf surfaces, suggesting competitive effects.
INTRODUCTION
Fungi have been increasingly recognized as an important group among invasive species (Desprez-Loustau et al. 2007, van der Putten et al. 2007, Mallon et al. 2015, Dickie et al.
2017), with some devastating consequences in the case of plant and animal pathogens (Fisher et al. 2012, Roy et al. 2017). Human-mediated transport has been identified as a major pathway for the introduction of non-native micro-organisms, e.g. forest pathogens (Liebhold et al. 2012). Microbial invasions are thus clear evidence that microbial cosmopolitanism (the absence of dispersal limitations) postulated in the "everything is everywhere" hypothesis is not the general rule (Green and Bohannan 2006, Martiny et al. 2006).
6 One acute problem in studying the spatial distribution of microbes, including fungi, which may explain apparent species cosmopolitanism, is the relatively low taxonomic resolution provided by morphological characters (Green and Bohannan 2006). For example, molecular analyses have shown that many 'morpho-species' of fungi hide a complex of genetically divergent species, called cryptic species, with follow-up studies providing evidence for some differentiation in their ecology and biogeography (Taylor et al. 2000). In particular, many plant diseases that were formerly believed to be caused by a single pathogen species were later shown to be due to a complex of multiple cryptic species (Fitt et al. 2006). For example, Eucalyptus leaf spot is associated with more than 60 species of Mycosphaerella (Crous and Groenewald 2005), and grapevine downy mildew is caused by five Plasmopara species in North America (Rouxel et al. 2013).
The existence of cryptic species may have important implications in the context of invasions, notably by causing so-called cryptic invasions (Geller et al. 2010). Invasions may be unrecognized due to the morphological similarity between native and introduced species (Geller 1999). In the recent European ash dieback , it took years to identify the exotic origin of the fungal pathogen (Hymenoscyphus fraxineus), closely related to a native non-pathogenic species, impeding the implementation of quarantine measures (Gross et al. 2014). Moreover, some invasions regarded as caused by a single species may hide several invasion events of different species (Mackie et al. 2012). For example, fungal chytridiomycosis has been identified as one of the major drivers of the decline of amphibians worldwide and was initially thought to be caused by the single species Batrachochytrium dendrobatidis (Fisher et al.
2009). However, a closely related species that preferentially infects salamanders has been recently detected and described (Martel et al. 2013). Cryptic species, although morphologically similar, often display some level of ecological divergence which could play a role in the invasion process. For example, differences in terms of environmental tolerance
may lead to some geographical segregation in introduced areas (Mackie et al. 2012). In the case of pathogens, as shown in the Batrachochytrium complex, adaptation to the plant or animal host is a major driver of ecological divergence and speciation (Le Gac et al. 2007, Giraud et al. 2010). Moreover, the ability to perform host shifts to non co-evolved hosts in the introduced area is a key process of pathogen invasions in wild communities (Woolhouse et al.
2005, Slippers et al. 2005). Differences in levels of host specialization within a species complex may cause different invasion patterns between cryptic species (Saleh et al. 2012).
Here we focus on oak powdery mildew, a foliar disease first recorded in the beginning of the 20th century in Europe, causing seedling mortality and tree decline, especially in the two most common European oaks, Quercus robur and Q. petraea (Marçais and Desprez-Loustau 2014). Recent molecular studies suggest a multiple invasion with three closely related fungal species in the Erysiphe genus putatively originating from Asia (Takamatsu et al. 2007, Mougou et al. 2008). The most common species, known as E. alphitoides, was described as a new species in 1912 (under the name of Microsphaera alphitoides) and regarded as the causal agent of the new invasion in Europe (Griffon and Maublanc 1912). The second species, E.
hypophylla, tentatively identified from slightly different symptoms and morphology, was assumed to cause an independent invasion, starting a few decades later from northern Europe (Roll-Hansen 1961). The third species, E. quercicola, was first described in Asia (Limkaisang et al. 2006, Takamatsu et al. 2007) and recently detected in France (Mougou et al. 2008) and later Spain (Desprez-Loustau et al. 2017). Its date of introduction in Europe remains unknown.
The overarching aim of this study is to understand the multi-scale spatial distribution of the cryptic species forming the invasive complex with the ultimate goal of understanding their invasive behaviour. Our working hypothesis is that species show some degree of niche
8 separation, which may translate into differences in their distribution across multiple spatial scales. More specifically, we address the following questions:
What is the spatial distribution of the three Erysiphe species, ranging from a large scale (i.e. the European continent) and local scale (i.e. stand level) to the micro-scale (i.e. leaf level)? At which spatial scales do the different species co-occur?
Do the fungal species differ in their biogeographical distribution at the continental scale due to differences in their climatic niche and their ability to attack different oak species?
At the stand scale, is the pathogen complex the same on trees of different developmental stages?
At the scale of single leaves, where direct and plant-mediated interactions between the pathogen species are expected to occur, do species segregate between and within leaf surfaces?
MATERIAL AND METHODS
Hierarchical sampling design
To assess the multi-scale distribution of the powdery mildew species complex, we sampled more than seven hundred powdery mildew-infected oak trees from across Europe using a series of sampling strategies (see below). The focal oak species was Quercus robur, which has the widest geographic range and spans all over Europe (http://www.euforgen.org/species/) but other oak species were also sampled (see below). The infected leaves were pressed dry immediately after field collection, sent to the laboratory and maintained at room temperature until the molecular analyses were performed. Unless otherwise stated, 6 mm diameter leaf discs were taken with a cork borer in sporulating lesions visible on the upper (adaxial) surface. Tools were sterilized between taking each leaf disc.
European scale sampling
The main sampling to assess the distribution of the cryptic powdery mildew species across Europe was completed during the growing seasons of 2014 - 2016 by collecting leaves from 20-80 oak trees, from one or several locations in each of 15 European countries and Turkey.
Additional data come from a previous sampling in France in 2007 (Mougou-Hamdane et al.
2010) and from previous collections in various countries (see Leaf scale sampling).Most samples were collected on Q. robur, with some of them identified as "Q. robur or Q.
petraea", since these two species are closely related and not easily distinguished. Other Quercus species with a more restricted range, especially Q. pyrenaica, Q. cerris, Q.
vulcanica, Q. pubescens, and Q. frainetto, were sampled in a limited number of countries (Desprez-Loustau et al. 2018). Individual samples for molecular analysis consisted of one leaf disc cut from one leaf per tree.
Stand scale sampling
In order to investigate whether the powdery mildew community is similar in mature trees and in seedlings growing under their canopy, we first sampled four mature trees growing approximately 50 m apart in Le bois des Sources in Cestas, France (lat. 44.76°, long. -0.71°), a location where E. alphitoides and E. quercicola were previously known to occur (Hamelin et al. 2016). For each mature tree, eight leaves from each of four branches were collected at circa 4 m high in the canopy using a pole pruner, in June 2014. The presence of Erysiphe species in seedlings was studied for two out of the four mature trees by collecting one infected leaf from each of 15 seedlings growing under the canopy of the mature tree at the same date.
In June 2015, we sampled again in the same location (Cestas) as well as in Laveyron (lat.
43.76°, long. -0.22°), another naturally regenerating stand in southwestern France where E.
alphitoides and E. quercicola were known to occur. Five trees were sampled in Cestas
10 (including the four sampled in 2014) and eleven in Laveyron, with five leaves sampled in the canopy of each mature tree and one leaf on each of five seedlings growing underneath each tree. Samples for molecular analyses consisted of one disc per leaf.
Leaf scale sampling
Three different samplings were performed at the leaf scale.
First, because the three species may differ in the location and intensity of sporulation, we tested whether visual sampling may introduce a bias in the detection of species.For this, we compared targeted vs. non targeted sampling in infected leaves, i.e. one leaf disc cut in a visibly sporulating lesion on the upper leaf surface and a second, "non-targeted", leaf disc taken on the same leaf, at the symmetric position from the main leaf vein (i.e., irrespective of any visual selection of lesions). Leaves were taken from the four trees used for canopy sampling in 2014 (total number of samples: 128 leaves and 256 leaf discs).
Second,differences in the occurrence of Erysiphe species between the upper and lower leaf sides were investigated on 74 samples collected from Armenia, Czech Republic, Germany, Iran, Israel, Lithuania, Slovakia, Switzerland, Hungary and the United Kingdom, as collected from 1969 to 2006. Mycelium and spores were scraped separately from lesions visible on lower (abaxial) or upper (adaxial) sides of leaves. If present, chasmothecia were also isolated separately for each leaf side.
Third, the fine scale distribution of E. alphitoides and E. quercicola within leaves, and their possible co-occurrence, was studied on leaves taken from different seedlings in Cestas and Laveyron forests in June 2015. The seven analyzed leaves (four from Cestas and three from Laveyron) showed extensive mycelial development on the upper surface (and not on the lower surface) and were known to harbor both E. alphitoides and E. quercicola based on previous analyses. Intensive sampling of the whole infected leaf surface was performed by taking many
small leaf discs (4 mm in diameter). Taking leaf discs with a cork borer was preferred over cutting the leaf into quadrats in order to avoid cross contamination while cutting.
Molecular identification
Identification of Erysiphe species was based on polymorphisms in the ITS (Internal Transcribed Spacer) sequences of ribosomal DNA, the universal barcode for fungi (Schoch et al. 2012).
In the case of leaf disc samples, total DNA was extracted from leaf samples using the Invisorb Spin Plant mini kit, according to manufacturer’s instructions (Stratec Molecular GmbH, Berlin). Amplification of the ITS1 region was performed using the ITS1-Fungi (Gardes and Bruns 1993) and o-micro-rev primers (Heuser and Zimmer 2002), as previously described (Desprez-Loustau et al. 2017). Overall, 94% of powdery mildew lesions produced the expected amplicon. Sequencing of the amplicon was done by Sanger technology (Genewiz, England). Sequences were aligned with BioEdit and Erysiphe species identified thanks to six fixed SNPs (Takamatsu et al. 2007, Mougou et al. 2008). Bad quality sequences or those that could not be assigned to one of the Erysiphe species were discarded (less than 3%).
In the case of leaf scrapings, DNA was extracted using the chelex method (Hirata and Takamatsu 1996). The rDNA ITS region including 5.8S rDNA was amplified by two sequential PCR reactions using partially nested primer sets according to the procedures of Takamatsu et al. (2008). The amplified ITS regions were digested with PvuII or Alw26 I (TaKaRa, Tokyo, Japan) according to the manufacturer’s recommendation. Digestion mixtures were prepared as 10 units of enzyme, 1.5 µl of enzyme buffer (supplied by the manufacturer), 7.5 µl of sterile distilled water and 5 µl of the PCR product per tube.
Digestions were run for two hours at 37˚C. Restricted DNA was then analyzed on 2% agarose
Kommentert [M1]: make a final check with table
12 gels in TBE buffer. Erysiphe species were identified according to band patterns. The ITS of E.
quercicola is cleaved by neither of the two enzymes. The ITS of E. alphitoides is not cleaved by Alw26 I, but is split into two bands (ca 450 and 100 bp length) by PvuII treatment. The ITS of E. hypophylla is split into two bands (ca 450 and 100 bp length) by PvuII treatment and also generates c. 300 and 250 bp bands by Alw26 I.
Climatic data
Climatic data were retrieved for each sampling location from WorldClim (Fick and Hijmans 2017) for the period 1970-2000, with a 2.5 minute resolution. Mean monthly temperature and rainfall were extracted and seasonal bioclimatic variables were computed by averaging temperature and summing rainfall for the spring (April to June), summer (July to September), autumn (October to December) and winter (January to March) over the 30-year period.
Statistical analyses
Statistical analyses were performed with the SAS 9.4 TS1M1 software (SAS Institute Inc., Cary, NC, USA). The probability of occurrence of the three Erysiphe species at the different spatial scales was studied by generalized linear models with binomial distribution and logit link. The effect of a set of predefined predictor variables on this probability was assessed by Wald Χ2 probabilities and odds ratio (OR) estimates. The OR quantifies the strength of the association between the dependent and independent variables, where OR>1 indicates a positive association and OR<1 indicates a negative association (Rita and Komonen 2008).
Analyses were either performed at the level of the individual sample (i.e. the leaf discs) or by grouping samples as explained below. We used a scale parameter (estimated by the ratio of
Pearson Chi-2 to degrees of freedom) to model the overdispersion in the data along with an AIC corrected for overdispersed data, i.e. QAIC (Burnam and Anderson 2003).
The analysis of the distribution of the three mildew species at the European scale (N = 689) was performed by pooling all samples within a country. To identify large scale geographic factors independently of host effects, only samples collected from Q. robur, Q. petraea and unidentified species within this complex were retained in the first statistical analysis (N = 498). For each Erysiphe species, a first model was run with proportion of positive samples per country explained by latitude and longitude (averaged among samples within a country). In order to investigate potential climatic effects underlying the geographic effects, the model was then run with the seasonal climatic variables (averaged within a country over 1970-2000) as predictor variables. We thereby assumed that the probability of detecting a given species at the national scale was mainly related to the establishment of a population of this species, itself depending on the long-term average climate. Due to the strong correlations existing between climatic variables, we first considered models with only one climatic variable. To determine the most influential climatic variables, we ranked models according to QAIC differences between the simple regressions (with a single climatic predictor) and the corresponding intercept models (with the same value of scale parameter for each pair of models, thus enabling QAIC comparisons). In a second stage, we ran models with all combinations of one temperature variable and one precipitation variable. The impact of host species on the occurrence of each of the fungal species was investigated separately for the four countries where sufficient samples had been collected from multiple oak species (at least 13 samples per oak species), i.e. Austria and Germany for Q. robur and Q. petraea, and Spain and Portugal for Q. robur and Q. pyrenaica. For each fungal species, we modelled the presence- absence on a single leaf disc cut from one leaf per tree as a function of the oak species.
14 At the stand scale, the proportion of positive samples for each Erysiphe species was computed for the 18 trees sampled over the two years and locations, separately for the canopy and the seedlings underneath the mature trees. We then modelled the proportion of leaves with each Erysiphe species as a function of tree developmental stage (mature vs. seedling), year (2014 vs. 2015) and location (Cestas vs. Laveyron).
Within leaves, the effect of targeted vs. non targeted sampling on detection of a given Erysiphe species was tested by modeling presence-absence as a function of sampling method for the 128 sampled leaves. To test for differences of occurrence of the fungal species on the two sides of a leaf, presence-absence of each species was modelled as a function of leaf side (upper vs. lower) for a total of 74 leaves.
To investigate the potential (positive or negative) interactions between species (E. alphitoides and E. quercicola) within a single leaf, we used the checkerboard score (Cscore) as an indicator of species segregation, as proposed by Stone and Roberts (1990). A low Cscore indicates a high randomness, i.e. a greater likelihood that the distribution of one species has not been directly affected by the presence of the other species.
First we calculated the Cscore from observed data:
) )(
( EA EAEQ EQ EAEQ
score N N N N
C
where NEA is the total number of leaf discs yielding E. alphitoides, NEQ is the total number of leaf discs yielding E. quercicola, NEA-EQ is the total number of leaf discs yielding both E.
alphitoides and E. quercicola. The Cscore was calculated for each sampled site independently and by pooling all samples.
Second we simulated the distribution of Cscore under the null hypothesis of randomness. Each Cscore was obtained by randomizing occurrences of E. alphitoides and E. quercicola among all
discs. To take into account potential differences in relative frequencies of Erysiphe species among sites (Cestas and Laveyron), the total number of each species occurrence was kept at the observed value in each site. This was repeated 1000 times.
Third we performed a two-tailed test by comparing the observed Cscore to the lower (αlow=0.025) and higher (αhigh=0.975) critical values of the null distribution of Cscore. The null hypothesis was rejected if the observed Cscore was lower or higher than the corresponding critical values, indicating more or less frequent co-occurrence than predicted by chance, respectively. These analyses were implemented in R software v. 3.3.3 (R Core Team 2017).
RESULTS
European scale
Clear geographical patterns emerged at the continental scale (Fig. 1). E. alphitoides was detected and was dominant in all European countries (Fig. 1). In contrast, the ranges of E.
quercicola and E. hypophylla were more limited, with almost non-overlapping distribution areas. E. quercicola was mostly restricted to the southern part of Europe and particularly common in the Iberian peninsula. It was also detected at very high frequency in Turkey and southwestern Asia, and was the sole species in the few samples from Israel and Iran. E.
hypophylla was most abundant in northern and central Europe, and almost absent from southwestern Europe, with the exception of a few leaves in Portugal and southern France. A significant proportion (22%) of samples (i.e. leaf discs) yielded mixtures of E. alphitoides with one of the two other species, whereas E. quercicola and E. hypophylla were only detected once on the same leaf (also with E. alphitoides).
16 Samples from Q. robur and/or Q. petraea covered a latitudinal gradient ranging from 40.29°
to 66.02° and a longitudinal gradient ranging from -8.37° to 25.61°. While we did not detect any significant effect of latitude or longitude on the distribution of the three Erysiphe species, climatic variables were shown to have a significant effect on the presence of E. hypophylla and E. quercicola (Table S1). Temperature, especially in fall and winter, had the most significant and important effects but in opposite directions for the two species. Erysiphe hypophylla was more common in a cold climate (OR around 0.75), while the reverse was true for E. quercicola, with OR around 1.5, meaning that the odds of finding E. quercicola are increased by c. 50% for each additional degree in mean winter temperature (Fig. 2; Table S1).
Among the precipitation variables, spring rainfall showed a significant effect, slightly negative for E. hypophylla and positive for E. quercicola. No better model with two variables was obtained for either species.
The probability of finding E. alphitoides, E. hypophylla or E. quercicola differed among oak species in areas where several species were present (Fig. 3). In Germany and Austria, E.
hypophylla was detected more frequently on Q. petraea than on Q. robur (ORs of 6.7 and 5.0 in 66 and 46 samples, respectively; P = 0.005 and 0.16, respectively). In Spain and Portugal, E. quercicola was detected much more frequently on Q. pyrenaica than on Q. robur (ORs of 6.3 and 9.6, in 51 and 85 samples, respectively; P = 0.03 and <0.0001, respectively). For E.
alphitoides, a host effect could only be tested for Portugal, with lower odds for Q. pyrenaica than for Q. robur (OR = 0.05, P = 0.0004).
Stand scale – Erysiphe species in the canopy of mature trees compared to seedlings growing under their canopies
In both sampling locations in southwestern France, almost all samples belonged to E.
alphitoides or E. quercicola. However, the relative proportion of species varied widely between tree canopies and seedlings, with E. quercicola much more frequently detected (alone or in mixture with E. alphitoides) in seedlings than in the canopy of overhanging mature trees (Fig. 4, OR = 6.2, Wald’s Χ2 = 23.6, P < 0.001). There was also significant variation in the frequency of E. quercicola among the two years (OR = 3.59 for 2015 vs.
2014), but no variation among the two locations.
Leaf scale – visual sampling effect and differences among the upper and lower leaf sides
There was a marked difference between discs taken from sporulating powdery mildew lesions on the upper leaf surface and discs taken randomly from the same leaves (Fig. 5). Although not as high as in the targeted sampling (91.3%), the detection rate of Erysiphe in leaf discs taken without targeting lesions reached 78.7%. The detection of E. hypophylla was higher in these non-targeted samples (i.e. irrespective of any visual selection) than in targeted samples (i.e. with visibly sporulating lesion), with 22% compared to 2%, respectively, and a very high odds ratio value of 15.8.
Infections and successful detection of Erysiphe spp. were more frequent for the upper leaf side than for the lower leaf side (71 vs. 39 out of a total of 74 leaves). Whereas E. alphitoides and E. quercicola were detected on both sides of the leaves, E. hypophylla was only detected from lesions sampled on the lower leaf surfaces, either as mycelium and conidia or chasmothecia (Fig. 6). Erysiphe alphitoides was present on the upper surface of most leaves where E. hypophylla was detected on the lower surface.
18 Within leaf co-occurrence patterns
Among the 177 leaf discs analyzed on the seven leaves, almost half (88, i.e. 48%) corresponded to a mixed infection by both E. alphitoides and E. quercicola (i.e. with both ITS detected, Fig. 7). In five out of the seven leaves, either E. alphitoides (for leaves C3, L7) or E.
quercicola (for leaves C1, C2 and L5) occupied the whole infected surface (i.e. was detected in all leaf discs, or almost for L5), with some of the discs also infected by the other species, i.e. showing mixed infection (Fig. 7). The leaves C4 and L6 were almost equally colonized by the two species, with most discs showing mixed infections and a few discs with only E.
alphitoides and others with only E. quercicola (Fig. 7).
The observed Cscores were 240, 816 and 1972, for Cestas, Laveyron and the whole set of leaves, respectively. These values were higher than the upper critical values for the Cscores generated under the null hypothesis of randomness (P = 0.012, <0.001 and <0.001, respectively), suggesting that Erysiphe species were more segregated than expected by chance (Fig. S1).
DISCUSSION
Multi-scale distribution patterns are less well documented for invasive species than for native species, although they can offer crucial insights into their invasive potential and ecology (Allen and Shea 2006, Brown et al. 2008). For micro-organisms, such information is even scarcer because difficulties in species identification have long hampered biogeographical studies (Taylor et al. 2000, Geller et al. 2010; but see Linde et al. 2002). Our investigation provides the distribution and co-occurrence data of three closely-related fungal invasive species, across spatial scales ranging from an oak leaf to the European continent. We show
that both environmental and host factors explain the different invasion patterns and the coexistence of the three species through niche segregation at the various spatial scales.
The continental scale: one broadly distributed species and two species with a restricted geographical range- a predominant effect of climate?
Only one of the three Erysiphe species, namely E. alphitoides, was found to be spread throughout Europe, thus showing high invasive success. Numerous studies have seeked to identify inherent traits or characteristics explaining species invasiveness (Sakai et al. 2001, Philibert et al. 2011), among which climate match has been shown to be a consistent predictor across biological groups (Hayes and Barry 2008). While the ecophysiological requirements for E. alphitoides are not precisely documented, its spatial distribution encompasses a wide range of geographic and climatic gradients, in agreement with previous studies along altitudinal gradients (Desprez-Loustau et al. 2010, Dantec et al. 2015), which suggests tolerance to a wide range of environmental conditions. In contrast, E. quercicola and E.
hypophylla showed more restricted and largely non-overlapping geographical distributions, and we found significant climatic variables explaining their distribution, which may reflect sensitivity to temperature, albeit in opposite direction for the two species. The most important variables for both species were temperature during the off-season (fall and winter), while there was no effect of the summer temperature, a pattern previously reported at a regional scale in France (Marçais et al. 2017). Hence, while many epidemiological studies focus on pathogen performance during the growing season, it may be the off-season that limits the distribution of plant pathogen species (e.g. Redondo et al. 2015), especially for obligate pathogens like powdery mildews, for which overwintering represents a critical stage (Tack and Laine 2014). One explanation for the success of E. hypophylla in colder climates may be the prolific production of efficient overwintering structures consituted by chasmothecia (the sexual fruiting bodies) (Roll-Hansen 1961, Sucharzewska 2009), in strong contrast to the
20 absence or limited number of such structures in E. quercicola (Feau et al. 2012). The importance of sexual overwintering structures in the pattern of invasions was also shown for Phytophthora infestans, the agent of potato late blight. This plant pathogen was restricted to European regions with oceanic climates in the first decades after its introduction, when it occurred as a single mating type, but it spread to colder climates when sexual reproduction became possible after the introduction of the second mating type (Yuen and Andersson 2013).
Whether the low production of chasmothecia is an inherent characteristic of E. quercicola, a consequence of bottleneck effects during the introduction into various regions (as occurred for P. infestans), or a secondary loss as observed in other species of invasive fungi (Gladieux et al. 2015) remains to be investigated. However, preliminary data suggests that E. quercicola is also characterized by a low production of chasmothecia in its putative native area in Asia.
Like in Europe, E. quercicola is reported to occur on average in warmer regions than E.
alphitoides in Japan (Takamatsu et al. 2007). It has been reported (in the asexual form) in other warm regions in the world, on different mediterranean and sub-tropical species (Takamatsu et al. 2007, Desprez-Loustau et al. 2017). The absence of E. hypophylla in warmer climates may relate to a high sensitivity to elevated temperature and possibly UV radiation (Willocquet et al. 1996), which may also explain its absence from the upper leaf surface.
An alternative reason for the widespread distribution of E. alphitoides, and, reciprocally, of more restricted distributions for E. quercicola and E. hypophylla, may be the timing of introduction of the three different species resulting in different residence time, a major factor in invasion success (Wilson et al. 2007). E. alphitoides, described at the time of the first devastating outbreaks of oak powdery mildew in Europe (Griffon and Maublanc 1912), may have been introduced first. E. hypophylla was not reported before the 1960s; our observations show that its distribution has not enlarged much ever since (Roll-Hansen 1961, Cruchet 1962,
Viennot-Bourgin 1968, Braun and Cook 2012). E. quercicola has been detected only recently in Europe (Mougou et al. 2008, Desprez-Loustau et al. 2017), following its description (Takamatsu et al. 2007). The findings of E. quercicola in this study represent the first records in Armenia, Bulgaria, Hungary, Iran, Israel, Portugal, Slovakia, Sweden and Turkey (Desprez-Loustau et al. 2018). However, the species was probably overlooked for a long time when no molecular tools were available for its identification, as suggested by this large distribution and its detection in specimens dating from the late 1960s or early 1970s (Desprez- Loustau et al. 2018). The total extent of suitable habitat is another important variable conditioning invasion success (Wilson et al. 2007), and this suitable habitat is represeneted by the distribution of susceptible hosts for obligate pathogens such as powdery mildews.
Variation in the susceptibility to a given pathogen within a host genus or family is a common observation for both plants and animals (Le Gac et al. 2007, Bancroft et al. 2011). We here confirmed differential susceptibility of oak species to the three Erysiphe species (Takamatsu et al. 2007, Marçais et al. 2017). The spread of E. alphitoides could have been favored by the high susceptibility of Q. robur, which is the most widespread oak species in Europe. In contrast, the other two Erysiphe species were more frequent on oak species with more restricted ranges, such as E. quercicola on Q. pyrenaica.
The stand scale: the impact of host developmental stage on the fungal community
Age-related differences in susceptibility or resistance to pathogens have commonly been reported for plants, and are associated with different resistance mechanisms acting in juvenile vs. mature stages (Develey-Rivière and Galiana 2007). Our results may thus suggest that E.
quercicola, mostly found in seedlings, is an agent of juvenile disease of oaks. This contrasts with E. alphitoides, which infects all stages. However, we emphasize that several factors other
22 than not only age-related resistance sensu stricto may explain the observed patterns, such as but also changes in plant growth patterns and architecture with aging may be involved, e.g.
through the amount and distribution of susceptible organs with aging (Calonnec et al. 2013), and microclimate, or vertical distribution of inoculum. For tree powdery mildews, overwintering within canopies relies on the presence of chasmothecia, which attach to cracks in the the bark surface (Takamatsu 2004). The absence (or low production) of chasmothecia in E. quercicola may thus explain its low frequency in tree canopies. In contrast, polycyclic growth in seedlings (i.e., occurrence of successive shoot flushes separated by the production of new buds within the same season) may favor bud infection and thereby the alternative mode of overwintering (Desprez-Loustau et al. 2014).
The leaf scale: spatial segregation vs. coexistence between and within leaf surfaces
Leaves are increasingly recognised as a complex and heterogenous environment providing habitat for diverse microbial communities (Vorholt 2012). Such within-leaf variation is caused by a diversity of factors, like topography, structural elements such as trichomes, resource aggregation and microclimate, wich operate at different spatial scales on foliar microbes and allow their co-existence (Esser et al. 2015). Our study confirmed that leaf side (upper vs. lower) is a first important factor structuring the Erysiphe complex on oak leaves. In agreement with previous observations (Takamatsu et al. 2007) and its Latin binomial (Cruchet 1962), E. hypophylla was only found on the lower leaf surfaces. This was in strong contrast to E. alphitoides and E. quercicola, which colonized both sides but were generally more abundant on the upper leaf surfaces. Comparison of fungal communities among upper and lower leaf surfaces are scarce, but Breeze and Dix (1981) suggested that upper leaf surfaces show higher spore deposition while lower leaf surfaces may provide more favourable
conditions for development due to lower levels of competition and protection from direct sunlight and rain washing. In our study, a lower competitive ability of E. hypophylla towards E. alphitoides was suggested by the results of the targeted vs. non-targeted sampling experiment, since E. hypophylla was much more frequent in randomly sampled discs than in those with profuse sporulation associated with E. alphitoides. These findings are in accordance with observations by early mycologists that E. hypophylla was often found in trees not infected by E. alphitoides and that its sporulation was much less abundant compared to E. alphitoides (Roll-Hansen 1961, Cruchet 1962). From a methodological point of view, our study demonstrates demonstrated that sampling the most obvious symptoms, as often done in plant pathology, can induce a bias when studying the frequency of cryptic species associated with the same disease. The relative frequency of E. hypophylla may thus have been underestimated in our study at the European scale with targeted sampling, even though this is unlikely to affect the general pattern of geographic distribution.
Spatial segregation within leaves at a smaller scale, on the same leaf side, was suggested to occur between E. alphitoides and E. quercicola, possibly reflecting competition effects.
Although less studied than in plants and animals, competition has been shown to occur within and between foliar pathogen species (Newton et al. 1997, Al-Naimi et al. 2005, Kozanitas et al. 2017). In the particular interaction between E. quercicola and E. alphitoides, priority effects could be strongly involved, as shown in other systems (Vannette and Fukami 2014), since the two species were shown to exhibit different temporal dynamics during the season, associated with the differences in overwintering modes (buds vs. chasmothecia) (Feau et al.
2012, Hamelin et al. 2016).
Further investigations at even finer spatial scales and integrating temporal dynamics, e.g.
with sequential and co-inoculation experiments, are now required to test several hypotheses arising from our study, related to competitive elucidate the processes underlying interactions
Formatert: Skrift: Kursiv
24 between Erysiphe species on oak leaves. including resource competition (Pozo et al. 2016).
(using sequential and co-inoculation experiments, as in Kozanitas et al. 2017). Likewise, experimental field and laboratory studies could tease apart the independent and interactive effects of host species, host age, inoculum density and microclimate on the differences in mildew communities observed between trees and seedlings.
CONCLUSION
Altogether, our results show that seemingly similar species exhibit different distributions across multiple spatial scales in their invaded area, even though all three Erysiphe species successfully completed a host shift to European oaks, especially Q. robur. Environmental and host factors (climate and micro-climate, host species, host development, leaf characteristics) acting at each spatial scale appeared were suggested as important drivers in the dynamics of this simple fungal community. Context-dependent mechanisms linked to the history of invasions, such as residence time, genetic drift or loss of sexual reproduction may also be involved. This study highlights the role of environmental and host heterogeneities at scales ranging from a single leaf to a continent as factors shaping invasion patterns in related pathogen species, through with niche differentiation allowing avoidance of limiting competition with other native or invasive species.
ACKNOWLEDGEMENTS
We thank Adrien Bolay, Caroline Gordon, Banga Grigaliunaite, Vasyl Heluta, Seyed Akbar Khodaparast, Gyongyver Nagy and Eleni Topalidou, for providing samples. Part of the
Formatert: Skrift: Kursiv
analyses were performed at the Genome Transcriptome Facility of Bordeaux (grants from the Conseil Régional d’Aquitaine n°20030304002FA and 20040305003FA, from the European Union FEDER n°2003227 and from Investissements d'Avenir ANR-10-EQPX-16-01).
Funding was obtained through the European BiodivERsA project "RESIPATH: Responses of European Forests and Society to Invasive Pathogens" co-financed by national funds (ANR- 13-EBID-0005-01 for France, Portuguese Science and Technology Foundation Ref.
BIODIVERSA/0002/2012). Other grants came from the French ANR Funfit project (ANR- 13-BSV7-0011), the Japan Society for the Promotion of Science Grant-in-Aid for Scientific Research (Nos. 16K07613 and16F16097), and the Slovak Research and Development Agency under the contract No. APVV-15-0210.
DATA AVAILABILITY
All data are available at Dryad Digital Repository data.inra.fr (DOI : ).
LITERATURE CITED
Al-Naimi, F. A., Garrett, K. A., and Bockus, W. W. 2005. Competition, facilitation, and niche differentiation in two foliar pathogens. Oecologia 143:449-457.
Allen, M. R., and Shea, K. 2006. Spatial segregation of congeneric invaders in central Pennsylvania, USA. Biological Invasions 8:509-521.
Bancroft, B. A., Han, B. A., Searle, C. L., Biga, L. M., Olson, D. H., Kats, L. B., Lawler, J. J and Blaustein, A. R. 2011. Species-level correlates of susceptibility to the pathogenic
Kommentert [M2]: to be done
26 amphibian fungus Batrachochytrium dendrobatidis in the United States. Biodiversity and Conservation 20:1911-1920.
Braun, U., and Cook, R.T. A. 2012. Taxonomic manual of the Erysiphales (powdery mildews).
CBS Biodiversity Series 11: 1–707.
Breeze, E. M., and Dix, N. J. 1981. Seasonal analysis of the fungal community on Acer platanoides leaves. Transactions of the British Mycological Society 77:321-328.
Brown, K. A., Spector, S., and Wu, W. 2008. Multi‐scale analysis of species introductions:
combining landscape and demographic models to improve management decisions about non‐native species. Journal of Applied Ecology 45:1639-1648.
Burnham, K.P., and Anderson, D.R. 2003. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach. Springer.
Calonnec, A., Burie, J. B., Langlais, M., Guyader, S., Saint-Jean, S., Sache, I., and Tivoli, B.
2013. Impacts of plant growth and architecture on pathogen processes and their consequences for epidemic behaviour. European Journal of Plant Pathology 135:479-497.
Crous, P. W., and Groenewald, J. Z. 2005. Hosts, species and genotypes: opinions versus data.
Australasian Plant Pathology, 34:463-470.
Cruchet, P. 1962. Présence en Suisse d’une Erysiphacée sur les chênes Microsphaera hypophylla Nevodovskij (Microsphaera silvatica Vlasov). Berichte der Schweizerischen Botanischen Gesellschaft 72:123-131.
Dantec, C. F., Ducasse, H., Capdevielle, X., Fabreguettes, O., Delzon, S., and Desprez‐Loustau, M. L. 2015. Escape of spring frost and disease through phenological variations in oak populations along elevation gradients. Journal of Ecology 103:1044-1056.
Desprez-Loustau, M. L., Robin, C., Buée, M., Courtecuisse, R., Garbaye, J., Suffert, F., Sache, I. and Rizzo, D. M. 2007. The fungal dimension of biological invasions. Trends in Ecology and Evolution 22:472-480.
Desprez-Loustau, M. L., Vitasse, Y., Delzon, S., Capdevielle, X., Marçais, B., and Kremer, A.
2010. Are plant pathogen populations adapted for encounter with their host? A case study of phenological synchrony between oak and an obligate fungal parasite along an altitudinal gradient. Journal of Evolutionary Biology 23:87-97.
Desprez-Loustau, M. L., Saint-Jean, G., Barrès, B., Dantec, C. F., and Dutech, C. 2014. Oak powdery mildew changes growth patterns in its host tree: host tolerance response and potential manipulation of host physiology by the parasite. Annals of Forest Science 71:563- 573.
Desprez-Loustau, M. L., Massot, M., Feau, N., Fort, T., de Vicente, A., Torés, J. A., and Ortuño, D. F. 2017. Further Support of Conspecificity of Oak and Mango Powdery Mildew and First Report of Erysiphe quercicola and Erysiphe alphitoides on Mango in Mainland Europe. Plant Disease 101(7):1086-1093.
Desprez-Loustau, M. L., Massot, M., Fort, T., Toïgo, M., Aday, G., Boberg, J., Braun, U., Capdevielle, X., Cech, T., Chandelier, A., Christova, P., Corcobado, T., Dogmus, T., Dutech, C., Fabreguettes, O., Faivre d'Arcier, J., Gross, A., Horta Jung, M., Iturrixa, E., Jung, T., , Junker, C., Kiss, L., Kostov, K., Lehtijarvi, A., Lyubenova, A., Marçais, B., Oliva, J., Oskay, F., Pastirčák, M., Pastirčáková, K., Piou, D., Saint-Jean, G., Sallafranque, A., Slavov, S., Stenlid, J., Talgø, V., Takamatsu, T., and Tack, A.J.M. 2018. Data from: From leaf to continent: the multi-scale distribution of an invasive cryptic pathogen complex on oak. Dryad Digital Repository.
28 Develey‐Rivière, M. P., and Galiana, E. 2007. Resistance to pathogens and host developmental
stage: a multifaceted relationship within the plant kingdom. New Phytologist 175:405-416.
Dickie, I. A., Bufford, J. L., Cobb, R. C., Desprez‐Loustau, M. L., Grelet, G., Hulme, P. E., Klironomos, J., Makiola, A., Nunez, M.A., Pringle, A., Thrall, P.H., Tourtellot, S.G., Waller, L., Williams, N. 2017. The emerging science of linked plant–fungal invasions. New Phytologist, 215(4), 1314-1332.
Esser, D. S., Leveau, J. H., Meyer, K. M., and Wiegand, K. 2015. Spatial scales of interactions among bacteria and between bacteria and the leaf surface. FEMS Microbiology Ecology 91(3):1-13.
Feau, N., Lauron-Moreau, A., Piou, D., Marçais, B., Dutech, C., and Desprez-Loustau, M. L.
2012. Niche partitioning of the genetic lineages of the oak powdery mildew complex. Fungal Ecology 5:154-162.
Fick, S.E. and Hijmans, R. J. 2017. Worldclim 2: New 1-km spatial resolution climate surfaces for global land areas. International Journal of Climatology 37(12):4302-4315.
Fisher, M. C., Garner, T. W., and Walker, S. F. 2009. Global emergence of Batrachochytrium dendrobatidis and amphibian chytridiomycosis in space, time, and host. Annual Review of Microbiology 63:291-310.
Fisher, M. C., Henk, D. A., Briggs, C. J., Brownstein, J. S., Madoff, L. C., McCraw, S. L., and Gurr, S. J. 2012. Emerging fungal threats to animal, plant and ecosystem health. Nature 484(7393):186-194.
Fitt, B. D., Huang, Y.-J., van den Bosch, F., and West, J. S. 2006. Coexistence of related pathogen species on arable crops in space and time. Annual Review of Phytopathology 44:163–182.
Gardes, M., and Bruns, T. D. 1993. ITS primers with enhanced specificity for basidiomycetes‐application to the identification of mycorrhizae and rusts. Molecular Ecology 2:113-118.
Geller, J. B. 1999. Decline of a native mussel masked by sibling species invasion. Conservation Biology 13:661-664.
Geller, J. B., Darling, J. A., and Carlton, J. T. 2010. Genetic perspectives on marine biological invasions. Annual Review of Marine Science 2:367-393.
Giraud, T., Gladieux, P., and Gavrilets, S. 2010. Linking the emergence of fungal plant diseases with ecological speciation. Trends in Ecology and Evolution 25:387-395.
Gladieux, P., Feurtey, A., Hood, M. E., Snirc, A., Clavel, J., Dutech, C., Roy, M., and Giraud, T. 2015. The population biology of fungal invasions. Molecular Ecology 24:1969-1986.
Green, J., and Bohannan, B. J. 2006. Spatial scaling of microbial biodiversity. Trends in Ecology and Evolution, 21:501-507.
Griffon, E., and Maublanc, A. 1912. Les Microsphaera des Chênes. Bulletin Trimestriel de la Société Mycologique de France 28:88–104.
Gross, A., Holdenrieder, O., Pautasso, M., Queloz, V., and Sieber, T. N. 2014. Hymenoscyphus pseudoalbidus, the causal agent of European ash dieback. Molecular Plant Pathology 15:5-21.
Hamelin, F. M., Bisson, A., Desprez‐Loustau, M. L., Fabre, F., and Mailleret, L. 2016.
Temporal niche differentiation of parasites sharing the same plant host: oak powdery mildew as a case study. Ecosphere 7(11) ):e01517.
Hayes, K. R., and Barry, S. C. 2008. Are there any consistent predictors of invasion success?
Biological Invasions 10:483-506.
30 Heuser, T., and Zimmer, W. 2002. Quantitative analysis of phytopathogenic ascomycota on leaves of pedunculate oaks (Quercus robur L.) by real-time PCR. FEMS Microbiology Letters 209:295-299.
Hirata, T., and Takamatsu, S. 1996. Nucleotide sequence diversity of rDNA internal transcribed spacers extracted from conidia and cleistothecia of several powdery mildew fungi.
Mycoscience 37:283-288.
Kozanitas, M., Osmundson, T. W., Linzer, R., and Garbelotto, M. (2017). Interspecific interactions between the Sudden Oak Death pathogen Phytophthora ramorum and two sympatric Phytophthora species in varying ecological conditions. Fungal Ecology 28: 86-96.
Le Gac, M., Hood, M. E., Fournier, E., and Giraud, T. 2007. Phylogenetic evidence of host‐specific cryptic species in the anther smut fungus. Evolution 61:15-26.
Liebhold, A. M., Brockerhoff, E. G., Garrett, L. J., Parke, J. L., and Britton, K. O. 2012. Live plant imports: the major pathway for forest insect and pathogen invasions of the US. Frontiers in Ecology and the Environment 10:135-143.
Limkaisang, S., Cunnington, J. H., Wui, L. K., Salleh, B., Sato, Y., Divarangkoon, R., Fangfuk W., To-anun, C., and Takamatsu, S. 2006. Molecular phylogenetic analyses reveal a close relationship between powdery mildew fungi on some tropical trees and Erysiphe alphitoides, an oak powdery mildew. Mycoscience 47:327-335.
Linde, C. C., Zhan, J., and McDonald, B. A. 2002. Population structure of Mycosphaerella graminicola: from lesions to continents. Phytopathology 92:946-955.
Mackie, J. A., Darling, J. A., and Geller, J. B. 2012. Ecology of cryptic invasions: latitudinal segregation among Watersipora (Bryozoa) species. Scientific Reports 2:871
Mallon, C. A., van Elsas, J. D., and Salles, J. F. 2015. Microbial invasions: the process, patterns, and mechanisms. Trends in Microbiology 23:719-729.
Marçais, B., and Desprez-Loustau, M. L. 2014. European oak powdery mildew: impact on trees, effects of environmental factors, and potential effects of climate change. Annals of Forest Science 71:633-642.
Marçais B., Piou D. , Dezette D., and Desprez-Loustau M.L. 2017. Can oak powdery mildew severity be explained by indirect effects of climate on the composition of the Erysiphe pathogenic complex? Phytopathology 107(5): 570-579.
Martel, A., Spitzen-van der Sluijs, A., Blooi, M., Bert, W., Ducatelle, R., Fisher, M. C., Woeltjes, A., Bosman, W., Chiers, K., Bossuyt, F., and Pasmans, F. 2013. Batrachochytrium salamandrivorans sp. nov. causes lethal chytridiomycosis in amphibians. Proceedings of the National Academy of Sciences 110:15325-15329.
Martiny, J. B. H., Bohannan, B. J., Brown, J. H., Colwell, R. K., Fuhrman, J. A., Green, J. L., Horner-Devine, M. C., Kane, M., Krumins, J. A., Kuske, C. R., Morin, P. J., Naeem, S., Ovreas, L., Reysenbach, A. L., Smith, V. H., and Staley, J. T. 2006. Microbial biogeography:
putting microorganisms on the map. Nature Reviews Microbiology 4:102-112.
Mougou, A., Dutech, C., and Desprez‐Loustau, M. L. 2008. New insights into the identity and origin of the causal agent of oak powdery mildew in Europe. Forest Pathology 38:275-287.
Mougou-Hamdane, A., Giresse, X., Dutech, C., and Desprez-Loustau, M. L. 2010. Spatial distribution of lineages of oak powdery mildew fungi in France, using quick molecular detection methods. Annals of Forest Science 67:212.
Newton, M. R., Kinkel, L. L., and Leonard, K. J. 1997. Competition and density-dependent fitness in a plant parasitic fungus. Ecology 78:1774-1784.
32 Philibert, A., Desprez‐Loustau, M. L., Fabre, B., Frey, P., Halkett, F., Husson, C., Lung- Escarmant, B., Marçais, B., Robin, C., Vacher, C., and Makowski, D. 2011. Predicting invasion success of forest pathogenic fungi from species traits. Journal of Applied Ecology 48:1381-1390.
Pozo, M. I., Herrera, C. M., Lachance, M. A., Verstrepen, K., Lievens, B., and Jacquemyn, H.
2016. Species coexistence in simple microbial communities: unravelling the phenotypic landscape of co‐occurring Metschnikowia species in floral nectar. Environmental Microbiology 18:850-1862.
R Core Team. 2017. R: A language and environment for statistical computing. R Foundation for
Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.
Rita, H., and Komonen, A. 2008. Odds ratio: an ecologically sound tool to compare proportions. Annales Zoologici Fennici 45:66-72.
Roll-Hansen, F. 1961. Microsphaera hypophylla Nevodovskij (M. silvatica Vlasov), an oak powdery mildew fungus. Meddelelser Fra Det Norske Skogforsøksvesen 17:37–61.
Redondo, M. A., Boberg, J., Olsson, C. H., & Oliva, J. (2015). Winter conditions correlate with Phytophthora alni subspecies distribution in Southern Sweden. Phytopathology, 105(9), 1191- 1197.
Rouxel, M., Mestre, P., Comont, G., Lehman, B. L., Schilder, A., and Delmotte, F. 2013.
Phylogenetic and experimental evidence for host‐specialized cryptic species in a biotrophic oomycete. New Phytologist, 197:251-263.
Roy, H. E., Hesketh, H., Purse, B. V., Eilenberg, J., Santini, A., Scalera, R., Stentiford, G. D., Adriaens, T., Bacela-Spychalska, K., Bass, D., Beckmann, K. M., Bessell, P., Bojko, J., Booy, O., Cardoso, A. C., Essl, F., Groom, Q., Harrower, C., Kleespies, R., Martinou, A. F., van Oers, M. M., Peeler, E. J., Pergl, J., Rabitsch, W., Roques, A., Schaffner, F., Schnidler, S., Schmidt, B. R., Schönrogge, K., Smith, J., Solarz, W., Stewart, A., Stroo, A., Tricarico, E., Turvey, K. M. A., Vannini, A., Vilà, M., Woodward, S., Wynns, A. A., and Dunn, A. M.
2017. Alien pathogens on the horizon: Opportunities for predicting their threat to wildlife.
Conservation Letters 10:477-484.
Sakai, A. K., Allendorf, F. W., Holt, J. S., Lodge, D. M., Molofsky, J., With, K. A., Baughman, S., Cabin, R. J., Cohen, J. E., Ellstrand, N. C., McCauley, D. E., O'Neil, P., Parker, I. M., Thompson, J. N., and Weller, S.G. 2001. The population biology of invasive species. Annual Review of Ecology and Systematics 32:305-332.
Saleh, D., Laarif, A., Clouet, C., and Gauthier, N. 2012. Spatial and host-plant partitioning between coexisting Bemisia tabaci cryptic species in Tunisia. Population Ecology 54:261- 274.
Schoch, C. L., Seifert, K. A., Huhndorf, S., Robert, V., Spouge, J. L., Levesque, C. A., Chen, W., and Fungal Barcoding Consortium. 2012. Nuclear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi. Proceedings of the National Academy of Sciences 109:6241-6246.
Slippers, B., Stenlid, J., and Wingfield, M. J. 2005. Emerging pathogens: fungal host jumps following anthropogenic introduction. Trends in Ecology and Evolution 20:420–424.
Stone, L., and Roberts, A. 1990. The checkerboard score and species distributions. Oecologia, 85: 74-79.
34 Sucharzewska, E. 2009. The development of Erysiphe alphitoides and E. hypophylla in the
urban environment. Acta Mycologica 44:109-123.
Tack, A. J., and Laine, A. L. 2014. Ecological and evolutionary implications of spatial heterogeneity during the off‐season for a wild plant pathogen. New Phytologist 202:297-308.
Takamatsu, S. 2004. Phylogeny and evolution of the powdery mildew fungi (Erysiphales, Ascomycota) inferred from nuclear ribosomal DNA sequences. Mycoscience 45:147-157.
Takamatsu, S., Braun, U., Limkaisang, S., Kom-Un, S., Sato, Y., and Cunnington, J. H. 2007.
Phylogeny and taxonomy of the oak powdery mildew Erysiphe alphitoides sensu lato.
Mycological Research 111:809-826.
Takamatsu, S., Havrylenko, M., Wolcan, S.M., Matsuda, S., and Niinomi, S. 2008. Molecular phylogeny and evolution of the genus Neoerysiphe (Erysiphaceae, Ascomycota). Mycological Research 112: 639–649.
Taylor, J. W., Jacobson, D. J., Kroken, S., Kasuga, T., Geiser, D. M., Hibbett, D. S., and Fisher, M. C. 2000. Phylogenetic species recognition and species concepts in fungi. Fungal Genetics and Biology 31:21-32.
Van der Putten, W. H., Klironomos, J. N., and Wardle, D. A. 2007. Microbial ecology of biological invasions. The ISME Journal 1:28-37.
Vannette, R. L., and Fukami, T. 2014. Historical contingency in species interactions: towards niche‐based predictions. Ecology Letters 17:115-124.
Viennot-Bourgin, G. 1968. Note sur les Erysiphacées. Bulletin Trimestriel de la Société Mycologique de France 84:117-118.
Vorholt, J. A. 2012. Microbial life in the phyllosphere. Nature Reviews Microbiology 10:828- 840.
Willocquet, L., Colombet, D., Rougier, M., Fargues, J., and Clerjeau, M. 1996. Effects of radiation, especially ultraviolet B, on conidial germination and mycelial growth of grape powdery mildew. European Journal of Plant Pathology 102:441-449.
Wilson, J. R., Richardson, D. M., Rouget, M., Procheş, Ş., Amis, M. A., Henderson, L., and Thuiller, W. 2007. Residence time and potential range: crucial considerations in modelling plant invasions. Diversity and Distributions 13:11-22.
Woolhouse, M. E., Haydon, D. T., and Antia, R. 2005. Emerging pathogens: the epidemiology and evolution of species jumps. Trends in Ecology and Evolution 20:238-244.
Yuen, J. E., and Andersson, B. 2013. What is the evidence for sexual reproduction of Phytophthora infestans in Europe? Plant Pathology 62:485-491.
36 Figure captions
Figure 1. Geographic distribution in Europe and neighbouring Asian countries of the three
Erysiphe species associated with oak powdery mildew. Circle sizes represent the sampling effort per country, ranging from 1-5 samples (smallest circles) to more than 40 samples (largest circles).
Figure 2. Relationship between winter temperature and probability of occurrence of Erysiphe
alphitoides, E. hypophylla and E. quercicola in Europe (the shaded area represents the 95%
confidence interval)
Figure 3. Distribution of the Erysiphe species on different oak species within the same countries. Shown are pooled data for Germany and Austria (top) and Spain and Portugal (bottom). EA stands for E. alphitoides, EQ for E. quercicola, EH for E. hypophylla.
Figure 4. The community composition of Erysiphe species in mature tree canopies versus seedlings growing under their canopies. EA stands for E. alphitoides, EQ for E. quercicola, EH for E. hypophylla.
Figure 5. Detection probability of Erysiphe species among pairs of discs taken on the same
leaves, either cut in a visibly sporulating lesion or in a non-targeted sample (i.e. cut without a priori information on the presence of a lesion). EA stands for E. alphitoides, EQ for E.
quercicola, EH for E. hypophylla.
Figure 6. The relative frequency of each of the three Erysiphe species on the adaxial (upper) and abaxial (lower) surfaces of 74 leaves. EA stands for E. alphitoides, EQ for E. quercicola, EH for E. hypophylla.
Figure 7. Distribution of Erysiphe alphitoides and E. quercicola at the leaf level, for leaves of Quercus robur collected in spring 2015 in Cestas (code 'C') and Laveyron (code 'L') in France. The diameter of leaf discs is 4 mm, with the exception of one or two larger leaf discs per leaf. Note that the scale varies among leaves.
Figure S1. Null distribution of Cscores as based on 1000 randomizations, over the whole set of leaves, or independently for each site. The vertical thick line represents the observed Cscore. Dashed lines represent the upper and lower critical values of the two-tailed test at a level of statistical significance of 5%.
38
Table S1. Results of generalized linear models with seasonal climatic variables (T = temperature, P = precipitation) tested separately as predictor variables of the presence of Erysiphe species. For each species, models are ranked by decreasing values of QAIC = (QAIC without predictor) - (QAIC with predictor). Predictors with a significant effect are indicated in bold.
Erysiphe species
Predictor variable
QAIC * Wald Chi2 P-value Odds Ratio (95% CI)
E. alphitoides Pspring 2 (104-102) 2.7 0.1 1.01 (0.99-1.02)
Twinter 1 (108-107) 2.7 0.1 1.18 (0.97-1.44)
Tfall 1 (107-106) 2.9 0.09 1.25 (0.97-1.63)
Pfall 1 (103-102) 2.2 0.14 1.01 (0.99-1.02)
Pwinter 0 (97-97) 1.3 0.26 1.00 (0.99-1.01)
Tspring 0 (95-95) 2 0.16 1.17 (0.94-1.44)
Tsummer -2 (84-86) 0.2 0.62 1.17 (0.80-1.71)
Psummer -2 (76-78) 0.5 0.46 1.01 (0.99-1.02)
E. quercicola Twinter 9 (72-63) 8.2 0.004 1.41 (1.12-1.79)
Tfall 8 (64-56) 6.8 0.009 1.68 (1.14-2.49)
Tspring 3 (61-58) 2.9 0.09 1.51 (0.94-2.42)
Pspring 1 (55-54) 4.1 0.04 1.01 (1.00-1.02)
Tsummer 0 (51-51) 1.9 0.17 1.46 (0.85-2.51)
Psummer 0 (44-44) 1.6 0.21 0.99 (0.97-1.01)
Pwinter -1 (45-46) 1.6 0.2 1.00 (0.99-1.01)
Pfall -1 (45-46) 0.8 0.36 1.00 (0.99-1.01)
E. hypophylla Twinter 13 (174-161) 11.4 0.0007 0.78 (0.67-0.90)
Tfall 11 (166-55) 11.3 0.0008 0.72 (0.60-0.87)
Pspring 11 (169-158) 9 0.003 0.99 (0.98-0.99)
Tspring 6 (136-130) 7.7 0.006 0.80 (0.68-0.94)
Pwinter 5 (141-136) 4.2 0.04 1.00 (0.99-1.00)
Tsummer 2 (119-117) 3.7 0.054 0.75 (0.57-1.00)
Pfall 1 (115-114) 1.9 0.17 1.00 (0.99-1.00)
Psummer -2 (106-108) 0.4 0.53 1.00 (0.99-1.02)
* QAIC values without predictor are different between models for a single species because a (different) scale parameter was included to account for overdispersion for each tested predictor
